2 ms·

Is this dataset correct? The accelerometer, gyroscope and proximity sensor has a ~9% false ratio here which I intepret as 9% laptops/desktops and 91% phones. T

by chx 7y ago

Is this dataset correct? The accelerometer, gyroscope and proximity sensor has a ~9% false ratio here which I intepret as 9% laptops/desktops and 91% phones. That's a bit too many phones. But use of Adblock is over 34%. Also, I see 15.36% Chrome PDF Plugin and 23.81% Native Client both which only exist on PCs and again doesn't really mesh with only 9% non-phone. What's going on...? Are there laptops with all three things listed...? Also, everything else puts Chrome market share above 60% these days, only 38% here.

XelNika 7y ago
> The accelerometer, gyroscope and proximity sensor has a ~9% false ratio here which I intepret as 9% laptops/desktops and 91% phones. Some of the stats are clearly broken. For the 7 day duration, those stats are closer to 95% false. For the 90 days duration, I see values like Platform 153.00% Linux x86_64 Screen height 253.53% 1080 Media devices 159.95% Timeout > Also, I see 15.36% Chrome PDF Plugin and 23.81% Native Client I don't see PDF reader stats at all. > Also, everything else puts Chrome market share above 60% these days, only 38% here. Easily explained by bias if it's even correct. The kinds of people who would try this service are also the kinds of people who would use a niche browser.